1. Field of the Invention
The invention relates to an apparatus and method for cooling an endless paper web after it exits a dryer of an offset printing system and, more particularly, relates to an apparatus and method for preventing the ink on the web from resoftening as the web traverses the first downstream chill roll and for preventing dew or frost formation on chilled components of the system when the system is taken off-line.
2. Discussion of the Related Art
In high speed offset printing processes, an endless paper web up to 72" wide is fed through an offset press at speeds up to 3000 feet per minute (34 mph), where it is printed on at least one and typically both sides with a thermoplastic ink. The printed web is then drawn through a dryer which dries the web by evaporating most of the solvents from the ink. It is important to note, however, that dryers are not intended to and do not evaporate all solvents from ink. Were this the case, the ink would become brittle and crack and fall off from the web, thus forming a nonusable product. Industry standard therefore is to evaporate only 75-95% (typically 80-90%) of the solvents from the ink, thereby improving the finished product.
In the industry standard process, the printed and dried web exits the dryer at a temperature of about 280.degree.-325.degree. F. and enters an insulated sheet metal housing or "smokehood" which traps solvent vapors which are emitted by the still-hot web. The hot web then travels alternately over and under a series of cooled chill rolls which cool the web to or near room temperature.
Referring now to FIG. 1, when a rapidly moving web W emerges from the dryer and smokehood and makes apparent contact with a first downstream chill roll C, a layer of air A is formed between the web W and the surface of the chill roll C, causing a web-to-chill roll surface clearance Ho on the order of 0.001"-0.002". It is with this air layer A and associated clearance that one aspect of the present invention is concerned, and the manner in which they are formed and the problems produced thereby will now be described.
A widely held misconception is that the air layer A is formed by a boundary layer of air following the moving web and/or rotating chill roll. While such boundary layers do exist, they form little or no part in the formation of the air layer A because the average speed of the air following the web or roll decreases rapidly with distance from the web and hence exhibits a sharply decreasing low pressure flow profile. The resulting boundary layer is easily eliminated. Indeed, it has been proven by calculations that, at web speeds of 2000 feet per minute, the boundary layer produced by the moving web W can be eliminated by increasing the tension T on the web by less than 3%. The boundary layer following the chill roll surface can be eliminated even more easily because it is much smaller than the boundary layer following the web surface due to the fact that it has a very short distance in which to form, i.e., only that portion of the chill roll which is not contacted by the moving web--typically less than 180.degree..
Others have theorized that the air layer A between the web W and the first downstream chill roll C is the result of centrifugal forces produced by the web W as it bends around the chill roll C. These forces were theorized to throw the web outwardly away from the chill roll surface. However, it has been mathematically proven that the centrifugal forces actually present in the typical chill roll stand are of the same magnitude of the boundary layer effect and can be accommodated just as easily as the boundary layer.
It has been discovered that the air layer A is actually formed by a hydrodynamic pumping action occurring as the web W approaches the chill roll C. Specifically, air following the converging surfaces of the web W and chill roll C is drawn into a wedge which rapidly decreases in thickness as the web W approaches the chill roll C. Drawing air into this area of rapidly decreasing cross section acts as a pump which compresses the air to form the very thin but relatively high pressure air layer A between the web W and the chill roll C. Unlike boundary layers which are at extremely low pressure and can be eliminated quite easily, this relatively high pressure air layer cannot be removed simply by increasing web tension a few percent. Indeed, web tension T could be increased to the web breaking point without sufficiently reducing the thickness of the air layer A. This problem is exacerbated by the fact that the pumping action produced by the converging web and chill roll surfaces increases with increased speeds, resulting in higher-pressure and thicker air layers at higher press speeds.
The presence of the air layer A between the web W and the chill roll C produces at least two problems. First, and probably most obvious, chill roll performance is degraded because the cool surface of the chill roll is not in intimate contact with the web, thus decreasing heat transfer efficiency. This decrease is rather dramatic because air is a relatively poor heat conductor. Accordingly, more and/or larger chill rolls are required for complete web cooling than would be required if the web W were always in intimate contact with the first downstream chill roll C.
A second and more insidious problem arising from the formation of an air layer A between the web W and the first downstream chill roll C is solvent condensation and resulting ink resoftening and "picking." As discussed above, the web W is still very hot as it approaches the chill roll C, and residual solvents continue to evaporate from the hot web surfaces as the rapidly moving web W makes apparent contact with the first chill roll C. The solvent vapors in the air layer A quickly condense and accumulate on the relatively cold outer peripheral surface of the chill roll C. The accumulated solvents are then reabsorbed by the surface S of the previously-dried web W, thus resoftening the ink. The resoftened ink is then offset or "picked" on the next downstream surface to be contacted by the surface S of the web, typically the third chill roll on the chill roll stand. The defects caused by this picking or offsetting are referred to as "condensate streaks."
Condensate streaking is exacerbated by the fact that it does not necessarily take place only on the first chill roll. As discussed above, the cooling efficiency of the first downstream chill roll C is decreased due to the insulating effect of the air layer A. This decreased efficiency may prevent the web from being cooled sufficiently on the first chill roll C to prevent further solvent condensation and the resulting condensate streaking on subsequent chill rolls.
The need thus has been established to eliminate the air layer formed between a web exiting a dryer and the first downstream chill roll over which the web travels, or to at least eliminate the condensate streaking resulting from this air layer. Many have recognized that the air layer could be eliminated by pressing the web into intimate contact with the chill roll. However, all previous efforts to this effect have proven unsuccessful.
For instance, U.S. Pat. No. 4,369,584 to Daane attempted to eliminate the problem of condensate accumulation by preventing the air layer between the web and the first downstream chill roll from ever forming by blowing high pressure air on the moving web from a nozzle or orifice. Intimate contact between the web and chill roll is never achieved with this device; the air gap is only reduced in thickness. The Daane '584 patent teaches that the nozzle outlet should be located within 0.5" of the line of tangency between the web and the chill roll to optimize jet utility. In actual practice, the orifice has to be installed several inches downstream of the tangent line so as to prevent the air from the orifice from causing the web to move or flutter as the web exits the smokehood. Except for relatively low press speeds (below 1500-1700 fpm), the Daane device did not achieve its goal. At press speeds of 1800 fpm and above (in common use today), excessive power is required to minimize the thickness of the air layer.
The Daane '584 patent also discusses the use of a mechanical nip roll to eliminate the air layer, but only as it applies to films or other webs that can be contacted without damage. There is no discussion of contacting a hot moving printed web without damaging the web or overheating the nip roll.
Others have recognized that the only practical way to achieve true intimate contact between the web and first downstream chill roll is to mechanically press the web directly onto the chill roll. One such device, disclosed in U.S. Pat. No. 3,442,211 to Beacham, pressed the web into intimate contact with the chill roll using a "squeegee-roll . . . coated or covered . . . with a layer of ink-resistant material such as a silicone compound or a synthetic plastic such as polytetrafluorethylene." The device failed to perform as predicted because the "squeegee-roll" surface absorbed heat from the endless web and quickly overheated. The overheated surface remelted the ink on the web, causing the ink to adhere to the hot surface and damage the printed product. Beacham attempted to overcome this deficiency by locating his nip roll at the point where the web was partially cooled and was leaving the chill roll rather than at the point of first web contact. Thus, nip rolls such as those proposed by Beacham damaged the printed web even worse than condensate streaking with no nip roll.
U.S. Pat. No. 4,476,636 to Gross describes a device which is designed to eliminate as much air as possible between the web and the chill roll surface. Gross states that the purpose of his invention is to achieve an air layer reduction rather than elimination of the air layer. Gross's use of a rubber covered "squeeze" roller applied directly to the surface of the chill roll has little or no effect on the cause of web flotation over the chill roll surface.
U.S. Pat. No. 5,111,595 to Bessinger is yet another attempt to overcome the problem of condensate formation on the first downstream chill roll causing ink resoftening. Bessinger speaks of a pressure roller to squeeze the web against the roller (chill roll). He admits to the impossibility of its use in practice because ". . . the web surface that faces away from the roller (chill roll) to be contacted cannot tolerate engagement by a solid object." Bessinger specifically wants to avoid contact with the outward web surface of a printed web traveling over a chill roll.
U.S. Pat. No. 5,184,555 to Quadracci is another attempt to solve the problem of condensate streaking at the chill roll. The Quadracci device attempts to cure the problem without direct contact to the chill roll or web. This device does not work in actual practice because it does nothing to eliminate the formation of condensate in the annular air space between the web and chill roll surface. So long as the printed web remains above a temperature of about 200.degree. F., solvent will continue to evaporate from the web and condense on the chill roll surface.
The Quadracci patent also describes the Baldwin chill roll wiper device in use today on many web offset press systems. This device helps alleviate the condensate problem, but does not eliminate it. The Baldwin device uses a porous, absorbent cloth material that makes contact with the first downstream chill roll surface in the area left between where the web leaves the chill roll and the point of first web contact with the chill roll. Examination has shown that this device removes a portion of the condensate, but not all of it.
Still another solution to the air layer problem was proposed in U.S. Pat. No. 5,121,560 to Daane (the Daane '560 patent), which sought acceptable ways to cool an elastomer-coated nip roll in order to overcome the problems produced by the Beacham process. The Daane '560 patent is assigned to the assignee of thee present application, and the inventors of the present invention were familiar with Daane's efforts. Daane points out the problems encountered when attempting to use an all-metal pressure roll or nip roll to press a web into intimate contact with a chill roll. Daane also discusses the problems involved when attempting to use an elastomeric pressure roll and states "the index of contact temperature preservation for an elastomeric pressure roll is very low and not effective for cooling the opposite side of the web which it contacts. In operation the elastomeric surface immediately becomes hot and does not cool and set the ink. Instead the pressure roll will pick and smear the ink to destroy the readability of the print . . . . Thus efficient cooling of the unset ink, especially on both sides of the web in high-speed printing, remains a significant unsolved problem."
Daane attempted to solve this problem by cooling the nip roll peripheral surface from the exterior. Accordingly, the Daane '560 patent proposed a technique of positioning a doctor roll adjacent the nip roll to provide a metered amount of coolant on the outer peripheral surface of the nip roll upstream from the web. A later commercial embodiment achieved the same effect using a doctor spray bar.
Two problems were associated with the externally cooled nip roll proposed by the Daane '560 patent. First, the coolant had to be applied to the undersurface of the web to prevent coolant from dripping onto the web and ruining the product. This technique required that the web contact the first downstream chill roll from below. Unfortunately, as many as 99% of existing chill roll stands contact the first downstream chill roll from above and thus are incompatible with Daane's technique. Second, it has been discovered that moisture is inevitably transferred to the web by the damp nip roll and that this moisture quickly accumulates on downstream chill roll surfaces. The web absorbs the accumulated moisture to the point that it becomes saturated and unusable. The process proposed by the Daane '560 patent thus solved the second problem produced by the Beacham technique only to produce a third problem which can be solved only at such great expense as to make the device unmarketable.
Another problem associated with many fluid-cooled rolls is that frost tends to build up on the outer surface of the nip roll if the system is temporarily taken off-line due to the condensation of moisture on the cold surface of the roll. Upon subsequent restart, this frost melts as the hot web traverses the roll, thereby wetting the web and causing ink streaking or even web breakage.